Trend following for small portfolios, part 2
• Never trust the name of an ETF
• Understand the details fully
• Adopt a clear and simple approach
Last week, I mentioned some of the problems with applying standard trend following on smaller accounts. The big problem is that futures contracts are large, which makes it difficult to put on enough positions to gain sufficient diversification. This week, I'll be talking about exchange traded funds (ETFs).
The original idea of ETFs is great but that doesn't mean that all ETFs are great. If you trade ETFs you need to be really sure of what the fund is doing. Never trust the name of an ETF.
The original ETFs were passive index investments, in which a computer program would invest the funds' assets to match the exact weights of an index. As between 80 to 90 percent of all mutual funds fail to return the same or more than their benchmark index (according to the SPIVA reports), these passive ETFs are clearly a rational alternative. With such instruments, you'll get the exact return of the index minus a minimum fee, which is a fraction of what the underperforming mutual funds charge.
The only problem is that the ETF industry has developed into a method to sell structured products to people who don't know what structured products are. By packaging structured products in an ETF wrapper and putting a simplified name on it, they are marketed to the unsuspecting retail audience. This, however, is not the topic of this particular article. What you need to know is that you should never trade an ETF that you don't fully understand the details of.
In general, stay away from leveraged ETFs, from short ETFs, from active ETFs and from synthetic ETFs. Go with straight up physically backed ETFs, with a clear, simple and well-described investment approach.
There are ETFs in many asset classes, covering equities, currencies, rates and commodities. The coverage is not as good as it is for futures but it's reasonable. For diversified trend followers, ETFs have potential but the expectations need to be set very differently than it does for futures traders. Apart from the more limited coverage, the big concern here is the cash utilisation. While futures traders need only pay a small amount in margin upfront, ETFs are cash instruments.
In this first test, I used 25 ETFs with broad asset class coverage, and applied a standard trend following futures model. The model trades long and short, entering into any emerging trend and puts a trailing stop behind. There has been no optimisations or changes to this default trend model. Nothing has been done to try to improve it for ETF use.
Simulated trend following returns - Diversified ETF portfolio
Source: Original research, simulation done in RightEdge
This shows that we can get a fairly good return profile for a long-term strategy. The gains compared to the drawdowns show a prudent, risk-averse strategy. Whereas the S&P 500 index, showing in blue, took quite a beating in 2008, our strategy kept on going. Of course, it also shows that we failed to achieve the kind of spectacular returns that futures trend following show over time. Even in 2008, which was a stellar year for trend following, we didn't show anything amazing.
The reason for this is the cash utilisation. Without the built in leverage of futures, it's not easy to get enough returns. Remember that for these kinds of strategies, we are taking on equal risk per position. That, of course, means that positions in low volatility instruments are much larger than those of the volatile ones. If a market tends to move by half a percent per day on average, you need to take on a larger position than for one that moves by two percent per day.
In the futures world this is not a big concern but with limited cash this can become a problem. The low volatility markets will quickly eat up all your cash reserves, leaving nothing for the rest. Running out of cash also introduces a random element into the systematic strategy. Your portfolio will be filled up with the first markets that give a signal, and ignoring the ones after.
Low volatility instruments can therefore be a problem for cash instrument strategies. The instruments with the smallest daily moves are the rates. Taking them away from the universe reduces diversification but increases profitability. You increase expected return, but also risk.
While ETFs are not optimal for diversified trend following, they can certainly be used. There are a few things we can do to squeeze out a little bit more return.
In the new simulation below, I've made a couple of basic adjustments. The model itself is the same except that I turned off short positions. Having short positions in makes for a smoother return profile over time but in this case it eats up cash without yielding much return. I also removed the rates markets all together, which is bad for diversification but makes more room for the higher volatility markets.
Now we end up with a strategy with decent long-term performance, and clearly beating the index. It does have sharper drawdowns at times than the strategy above but that's the cost of doing business here.
Simulated trend following returns - Long only, no rates
Source: Original research, simulation done in RightEdge
Now we're up to a compounded return of around 10 percent, which is far higher than the global stock markets have managed to produce. And that's without any special adjustments to the trading rules.
Ten percent is not enough? It's a lot more than most people have been able to make over the years. However, it's a fair point and I shall be looking at ways to get the returns higher in another article.
Andreas F. Clenow is a hedge fund manager and principal at ACIES Asset Management. He is also the author of the best-selling book Following the Trend (Wiley 2012).